BTC $80,609.60 +0.62%
ETH $2,341.85 +0.77%
BNB $741.91 -0.22%
XRP $1.33 +1.24%
SOL $101.11 +0.35%
TRX $0.3310 +0.64%
DOGE $0.0787 +1.31%
ADA $0.2053 -0.33%
BCH $294.76 +4.64%
LINK $11.56 +1.68%
HYPE $81.43 +3.57%
AAVE $133.28 -1.41%
SUI $0.9290 +1.24%
XLM $0.1702 +0.63%
ZEC $1,298.10 +5.36%
AAPL $346.92 +0.89%
AMZN $254.51 +1.83%
GOOGL $326.82 +2.45%
MSFT $532.50 -0.08%
META $622.00 -1.98%
NVDA $218.13 +0.44%
TSLA $407.73 +3.29%
SNDK $1,326.43 +3.15%
INTC $94.30 +2.42%
SPCX $149.63 +0.96%
MU $947.00 +3.00%
AMD $563.66 +4.65%
BTC $80,609.60 +0.62%
ETH $2,341.85 +0.77%
BNB $741.91 -0.22%
XRP $1.33 +1.24%
SOL $101.11 +0.35%
TRX $0.3310 +0.64%
DOGE $0.0787 +1.31%
ADA $0.2053 -0.33%
BCH $294.76 +4.64%
LINK $11.56 +1.68%
HYPE $81.43 +3.57%
AAVE $133.28 -1.41%
SUI $0.9290 +1.24%
XLM $0.1702 +0.63%
ZEC $1,298.10 +5.36%
AAPL $346.92 +0.89%
AMZN $254.51 +1.83%
GOOGL $326.82 +2.45%
MSFT $532.50 -0.08%
META $622.00 -1.98%
NVDA $218.13 +0.44%
TSLA $407.73 +3.29%
SNDK $1,326.43 +3.15%
INTC $94.30 +2.42%
SPCX $149.63 +0.96%
MU $947.00 +3.00%
AMD $563.66 +4.65%

ssi

All
Article
Flash

Slow Fog: Udesk customer service script was implanted with malicious code that can steal user session data and replace recharge addresses

According to Slow Fog's disclosure, its security personnel discovered that the script udeskApi.2026-10-16.cached.js related to the Udesk customer service system was implanted with a malicious loader during routine threat intelligence checks. This loader hides the address through hexadecimal encoding and asynchronously loads the remote script deeps.js (sourced from udeskqlobal[.]com, note that the domain name starts with the letter q). This script reads accessible Cookies, localStorage, and sessionStorage on the page, intercepts form and login/authentication requests, and sends data to udeskqlobal[.]com/c.php. Attackers may use this to impersonate sessions or take over accounts. This data collection behavior is not restricted by the target site's rules, and unmatched pages may also be affected.According to Slow Fog's analysis, this script also targets three types of sites: mgbx[.]com, bikingex[.], and bitxex[.]io, scanning the recharge page every 200 milliseconds and replacing identified recharge addresses with preset BTC, EVM, TRON, Solana, LTC, and DOGE addresses. At the same time, the script manipulates the clipboard API, replaces copy buttons, and intercepts copy events, making users see and copy tampered addresses. If users transfer funds to this address, the money will flow into the attacker's address.

first_img South Korean police arrested 7 people in the first large-scale AI-assisted hacking case

Seoul police arrested 7 suspects accused of using generative AI to launch hacking attacks. The Seoul Metropolitan Police Agency stated that about 90 organizations were involved, collectively stealing 8.2 million records, marking the first large-scale arrest of suspects in AI-assisted hacking attacks in South Korea. The targeted victims included hospitals, universities, private tutoring centers, golf courses, and subsidiaries of large corporate groups, all of which held a significant amount of personal information, including names, contact details, and records of ordinary citizens. One funeral service company lost 1.22 million records, and reports indicated that the tutoring center's data also contained information on minors.According to local media reports, the main suspect, 32 years old, studied arts and sports and has no IT background. He told investigators that he self-taught hacking techniques through YouTube and online blogs, while police pointed out that he had worked in a security company conducting simulated attacks from 2019 to 2022. He initially used website vulnerability scanning tools and later introduced AI coding proxies and an automated penetration testing system called Artex at the end of August. Police stated that he hacked into 60 companies over a span of 44 days from August 31 to October 13.The entire attack operation lasted about 13 months, from September 2025 to this October, and the case was filed in June after a victim reported it. Police indicated that such a large-scale attack previously required a hacker group to accomplish, but AI now allows one person to achieve this in a short time. The suspect sent emails to 38 victims, collectively extorting 360 million Korean won, approximately 272,000 US dollars, but the only confirmed payment was 300,000 won, about 210 dollars. The main suspect was detained and transferred to prosecutors last Friday, facing charges of violating South Korea's Information and Communication Network Act, as well as extortion and attempted extortion. Police will continue to investigate the other 6 accomplices and confirm more victims, with 15 institutions already reporting damages.

first_img Google DeepMind released XYEval, assessing whether agents comply with misleading suggestions

Google DeepMind researchers Zhengxuan Wu, Yuxuan Li, Oyvind Tafjord, and Been Kim released a preprint paper titled "XYEval: Agents say yes to bad advice," proposing a meta-evaluation framework XYEval to assess the performance of agents when faced with misleading suggestions from users. The paper extends the "XY problem" to the agent scenario, where the user inquires about their attempted solution X rather than the actual problem Y.XYEval retains the original task environment and verifiable answers, only adding seemingly reasonable but misleading user suggestions in the instructions. The misleading suggestions are generated by a generator after seeing the complete task and are appended in the form of confident suggestions rather than hard requirements, allowing the original task to remain solvable. The evaluation covers six benchmarks: τ2-bench, SWE-bench Verified, SWE-bench Pro, Terminal-Bench, Humanity's Last Exam, and MCP-Atlas, along with five models, where the misleading suggestions for τ2-bench are generated based on rules set by human experts.The paper states that the performance significantly declines after the XY mutation, with the relative drop reaching up to 46.7%. On τ2-bench, when faced with users who require a detailed explanation before accepting a better solution, the performance drops even more. Encouraging models to recognize the XY problem through simple system instructions only partially alleviates the issue. The related code is published in the GitHub repository google-deepmind/xyeval.

The Smarter Web Company increased its holdings by 20 BTC, Futu Securities (Hong Kong) opened BNB deposits and withdrawals for professional investors

According to BBX data, last weekend, publicly listed companies and brokerages worldwide disclosed the latest developments in digital asset treasury allocation and trading channel services, with the core information as follows:The Smarter Web Company increased its holdings by 20 bitcoins: The London-listed technology company The Smarter Web Company officially announced that it recently made another asset allocation, increasing its holdings by 20 bitcoins. After this purchase, the company's total bitcoin holdings on its balance sheet have steadily risen to 2,850 BTC.Marinade released Q3 financial report, SOL staking volume increased significantly by 21.2%: The well-known Solana staking ecosystem project Marinade released its third-quarter financial report. The data shows that its SOL staking volume achieved significant growth, increasing by 21.2% to 7.94 million SOL. The company generated approximately $1 million in revenue for the quarter, demonstrating a robust cash flow in the on-chain staking ecosystem.Futu Securities (Hong Kong) launched BNB deposit and withdrawal services for qualified professional investors: Futu Holdings' Futu Securities International (Hong Kong) Limited officially announced the full launch of BNB deposit and withdrawal services for qualified professional investors (PI) in Hong Kong. Qualified investors can now use the BNB Smart Chain (BEP20) network to deposit funds from self-custody wallets to Futu crypto accounts or withdraw to verified whitelist addresses. Previously, Futu's wholly-owned licensed virtual asset trading platform in Hong Kong, PantherTrade, had already launched BNB trading pairs.
app_icon
ChainCatcher Building the Web3 world with innovations.